Video advertisements are among the most advanced, complicated, and expensive, forms of advertising content. Beyond the costs to produce video content itself, the expense of delivering video content over the broadcast and cable networks remains considerable, in part because television (TV) slots are premium advertising space in today's economy. Furthermore, TV is no longer a monolithic segment of the media market, and viewing data for TV content is tracked in a number of different ways. Consumers can now spread their viewing of video content, particularly premium content, across traditional TV, DVR, and a menagerie of over-the-top and on-demand video services viewed across smart TVs, gaming consoles, and mobile devices, as well as traditional TVs.
In short, TV viewing is transforming to digitally distributed viewing, as audiences watch proportionately less live broadcasting and more in a video on demand (VOD) or streaming video format.
Adding online consumption to the list of options available to any given consumer, only lends greater complexity to the process of coordinating delivery of video adverts to a relevant segment of the public. This complexity means that the task of optimizing delivery of advertising content today far exceeds what has traditionally been necessary, and what has previously been within the capability of experienced persons. The data needed to fully understand a given consumer is fragmented as each individual and household views more and more media in a disparate fashion by accessing a network of devices. In short, today's complexities require specifically tailored technological solutions, and take the decision making out of the hands of skilled people by utilizing computer methods that are able to handle a large number of factors, and at a speed, that humans could not possibly cope with. For example, human analysts have guided, and in some situations continue to guide, the selection of advertising inventory based on, for example, spreadsheets and other static data management tools designed for the desktop. But this results in low selection efficiency and delays in responding to market trends. Such methods are also incapable of quickly and accurately integrating information about how consumers behave across all of their devices.
Consequently, there are many important considerations that influence an advertiser's selection of advertising inventories and the type of content to deliver. The considerations include factors such as: time of day the advertisement will play, desired number of impressions, type of audience the advertiser wishes to reach, and the price of the advertising time slot.
Nevertheless, advertisers are heavily dependent on information they receive from media conduits and panel based data providers for assistance in deciding where and when content should be delivered, as well as assessing effectiveness of that delivery when making decisions on subsequent strategies. The decisions of how to deliver content, and what form that content takes, are particularly influenced by information about the viewing data made available by the content providers. For example, content providers can inform advertisers which demographic groups are likely viewers of a given program, according to time of day and program content. However, today's rich media environment demands attention to more factors when deciding when to deliver advertising content and to which devices.
Furthermore, in the context of today's advertising, it is both important but difficult to be nimble and flexible in content delivery: an advertiser wants to be able to react quickly to changes in market conditions or to specific occurrences such as a news development or a big sporting event; an advertiser also wants to act on an appreciation that an initial strategy is not optimal, as well as to capitalize on the consumer's access to many different viewing platforms.
Additionally, media conduits are effectively siloed and produce an environment in which it is not possible to coordinate an advertising campaign across both television media and digital video platforms at the same time. Often, advertisers deploy different teams and tools for each conduit. For example, Internet companies Google and Facebook are considered as media conduits because they have their own platforms for broadcasting content to a dedicated population of consumers. Each such company limits exchange of data to within their own properties. Similarly, an advertiser cannot easily coordinate delivery of content between, a social media network such as Facebook, and a TV content provider such as DirecTV. Consequently, many advertising agencies divide their campaign budgets between TV and online delivery.
However, combining data across content owners, devices, and media formats today is costly, impracticable, and in some cases unlawful due to government privacy regulations. Given these restrictions, advertising inventory purchasing today relies largely on limited data models that imperfectly pair advertising to market segment targets.
Therefore, it has not been possible with today's tools to track exactly which person has watched a particular advertisement because it is not possible to aggregate information from all the available media conduits on which that individual might have viewed content.
Most TV advertising buying decisions are based off of panel data for targeted audiences (such as that provided by Nielsen), which by definition involves polling a fixed group of consumers that have been selected by the ratings companies to be representative of the population at large. Human panelists report the TV content they have seen. Then statistical models on sample data and reporting metrics are produced to extrapolate to regional or national viewership trends and behaviors from the models. For example, advertisers receive data from TV panel companies (ratings agencies), and use the information to decide how they are going to design and implement an advertising campaign. Separately, online viewership data can be tracked by content providers such as Comcast, Netflix and YouTube. So, alternatively, the advertisers will receive data from online panels such as Comscore, Nielsen and Kantar, which track where the audience is online. Cable operators also sell their own viewership data from their subscribers. But these various sources are analyzed by advertisers and brand managers independently of one another. An additional drawback of panel-based models is their reliance on a fixed and relatively small number of parameters to characterize the viewing public.
While informational tools today are able to quantify viewer participation by calculating views per media device or provider, and infer, based on available census data, which types of individuals are likely to view an advertisement, the ability to aggregate exact viewer behavior across multiple media conduits has not been possible to do with useful accuracy or speed. As such, advertisers anticipate that in order to reach the desired audience, they will need to repeatedly play the same short clip either across many media conduits or target a selection of popular media conduits for multiple successive broadcasts of the same content or non-redundant versions of it. But the challenge of anticipating which viewers will actually view the content a certain number of times (frequency) remains.
Assessing whether a user has viewed TV delivered content has historically been challenging because it is difficult to establish whether a person actually watched the show or segment as it was being broadcast.
The advent of “Smart TV's” (also known as connected TV's) such as those manufactured by OEMs such as Samsung, LG, and Vizio has, however, provided more reliable means of measuring this data because Smart TV's allow consumers to opt into online connectivity of their TV sets. The TV's are connected to the internet as well as a feed from, say, a cable company, and so they send a data feed of programs being viewed on a particular TV, in real-time via the internet.
Data from Smart-TV's can be used to produce measurements that are at least as informative as those relied on by Nielsen, and offer the prospect of being superior for a number of reasons: the data that can be received from a SmartTV is richer than a simple yes/no response to whether a given viewer watched a particular program; there are many more SmartTV's in circulation than even the largest panels deployed by ratings companies, and that number continues to increase over time; and SmartTV data can potentially be linked to other data about a given consumer. This means that it no longer makes sense to rely on an old-fashioned technique that relies on a panel of consumers to validate a model.
Nevertheless, online media distributors such as Google don't have data from SmartTV's. Given this, the state of the art in advertising strategies differs across different media. For example, digital advertising is able to target based on known online behaviors, whereas TV advertising strategy is based on census data and is focused on reaching particular demographics. Furthermore, although data driven and automated TV advertising known as Programmatic TV (PTV) advertisement spending now constitutes 7% of all TV advertising spend, and is doubling every year, there remain differences between the feedback time for purchasing decisions in the TV realm from those relating to online content. Thus, it would additionally be useful to have in place bidding methods for purchasing advertising inventory that are tailored to programmatic TV content so that advertisers can reliably reach the growing segment of the population for which data is known via programmatic TV viewing.
Given the absence of more complex consumer classifications, predicting the behavior of consumers using available statistical methods can be ineffective. Consumer classification data is core to the decision making process for advertisers and brand managers when purchasing advertising inventory.
Pairing viewer TV behavior to online viewership data is not currently practiced. Panel data alone hasn't allowed analysts the ability to gather aggregated consumer behavior, and analyze it on a person-to-person basis. Today, the availability of unique user behavior can be better understood if aggregated across devices and media providers. However, privacy laws restrict the ability to commercialize, collect or share personally identifying user data collected from media, device and service operators. It is also unlawful in some instances to aggregate data that is protected by privacy laws. Thus, the intelligence that can be gained from that data must not include information relating back to consumer identities due to the risk of unlawful disclosure. Using identification hashing protects user privacy while permitting data aggregation, but current methods do not provide advertising buying platforms the ability to quickly and securely hash identities effectively.
There is therefore a need for a system to offer a unified purchasing experience for video advertising inventory and one that can reliably target relevant populations of consumers.
The discussion of the background herein is included to explain the context of the technology. This is not to be taken as an admission that any of the material referred to was published, known, or part of the common general knowledge as at the priority date of any of the claims found appended hereto.
Throughout the description and claims of the application the word “comprise” and variations thereof, such as “comprising” and “comprises”, is not intended to exclude other additives, components, integers or steps.